Who Benefits the most from Generative UI

Feb 25, 2024

Generative AI in UI design goes beyond automation; it's about augmentation and innovation. Generative AI platforms can quickly create wireframes and code. Most current tools also enable the export of these AI designs to popular traditional design tools, such as Figma. But who truly benefits most from these generative UI platforms?

This list will discuss some of the key players in this market and aim to establish a unified standard for benchmarking these systems.


V0.dev, currently supported by Vercel, is unique on this list as it only supports code generation. The platform uses GenAI processes to produce custom React components, styled using Shadcn for now.

After the code is produced from the initial input prompt, users can refine the generated React code using a chat-like interface. However, unlike Galileo, only one design is generated per prompt, not multiple variants.

V0.dev is highly suitable for teams and individuals familiar with their existing tech stack. It excels at creating intuitive designs. Yet, by choosing to use React+Shadcn, users are somewhat directed towards using the same frameworks.

An open-source version of this framework is available at: https://github.com/raidendotai/openv0.

Galileo AI

Galileo AI describes itself as the ChatGPT for interface design, offering an intuitive chat-based interface where users can generate multiple variants of the same design, similar to Dall-E.

In addition, users can further iterate and make changes to these designs using usual prompt-based strategies to achieve their final design. I found that the iteration process on this platform is much easier and more frictionless than on other platforms on this list.

Furthermore, as a non-designer (i.e., someone largely unfamiliar with Figma and similar tools), this interface felt more intuitive for both creating and iterating on AI-generated designs.

Lastly, for users on the free plan, up to 3 AI-generated designs can be exported to Figma for further editing, but the platform does not currently support code generation.


Uizard (uizard.com) recently introduced their AutoDesigner feature, which utilizes advancements in the GenAI field to enhance their in-house design editor.

The AutoDesigner, which caters to mobile, tablet, and web designs, accepts the following inputs:

  • A text-based prompt outlining the semantics or structure of your desired design

  • A text-based prompt describing your style or aesthetic

  • Alternatively, a screenshot of an existing UI that can be replicated into editable design components

The main prompt has a 300 character limit. The platform operates within a closed ecosystem - users are encouraged to refine the AI-generated designs within Uizard and they currently can't be exported to platforms like Figma.

Uizard also provides a Text2Code feature, allowing the export of AI-generated components into CSS-styled React components. However, this feature requires a paid subscription.


The last platform on today's list is called Visily.

Visily, a design platform, refers to AI as the foundation of its software, offering several AI features to complement its in-house design editor, similar to Figma.

As of now, the AI features on Visily's platform, in the context of generating UIs/wireframes, support only screenshot to design. The platform, similar to Uizard, takes in a picture of an existing interface and produces an interactive and editable wireframe.

These wireframes can be edited using the in-app editor (similar to Figma), but can also be exported as Figma components to be pasted into a Figma workspace.

In addition to this feature, the platform also offers multiple AI features such as:

  • color palette extraction

  • sketch-to-design, which is a feature still in beta. When we tested this out, it was shaky at best.


While Generative AI has extreme potential to completely revamp the design space, both in terms of reducing human effort and supplementing human creativity, there seem to be multiple ways platforms have tried to approach the problem.

These range from:

  1. Img2Code: Taking in a screenshot of a design and outputting a front-end implementation.

  2. Img2Design: Taking in a screenshot of a design and outputting a high-fidelity editable wireframe.

  3. Text2Code: Taking in a prompt description and outputting a front-end implementation.

  4. Text2Design: Taking in a prompt description and outputting a high-fidelity editable wireframe.

  5. Iterative/One-time: Allowing users to iterate upon their designs by re-prompting the AI or not.

Taking a broad view, it is also important to consider how AI will affect the incumbents in this space, most notably Figma. Figma has already rolled out AI features to its jamboard focused product, FigJam, and recently acquired Diagram, a small team "reimagining UI design in the era of generative AI". Like the mobile and the web before it, the rise of AI represents a paradigm shift that's changing the way designs are conceptualised, created, and shipped.

All in all, as AI reshapes the design landscape, the challenge will be to harness its power to complement human creativity, rather than replace it. This could lead to a future where AI acts as a co-creator, enabling designers to push the boundaries of what's possible and bring more innovative, inclusive, and user-centric designs to life. The journey ahead is as exciting as it is uncertain, but one thing is clear: the fusion of AI with design is not just a trend, but a new chapter in the evolution of digital creativity.